Models / calibrated-classifier

Undercut

P(pitting now puts the driver ahead of the car in front once it covers), when within 5s — plus WHEN to strike: the recommended undercut lap and its success probability.

calibrated-classifieranalytics.strategy

How it works

A calibrated model of the covered undercut duel: the attacker stops, the car ahead responds, and the outcome is read once both have stopped. The success probability depends on the gap to the car ahead and marginalizes the rival's reaction over its empirical response distribution. The timing call then compares pitting now with waiting. Served probabilities are the calibrated marginal at the current state and at the recommended strike state.

Inputs

  • gap to car ahead
  • car ahead's tyre age
  • rival reaction distribution
  • laps remaining

Output

drivers[].threat.undercut.pSuccessCalibrated probability that striking now ends ahead of the covering car.
drivers[].threat.undercut.viableWhether the window is open at all.
drivers[].threat.undercut.callRecommended strike lap, pNow, and pAtRecommended.

Calibration

Calibrated against historical undercut outcomes. See the calibration reference for the shared pipeline.

← All models